#computeruse

2025-10-27

Shortcuts 개발팀이 만든 Mac AI ‘Sky’, OpenAI 품에: 에이전트 시대 본격화

Apple Shortcuts 원조 개발팀이 만든 Mac AI 'Sky'를 OpenAI가 인수했다. 화면을 보고 직접 행동하는 에이전트 AI 시대가 본격화되며, OpenAI·Anthropic·Google의 3파전이 시작됐다.

aisparkup.com/posts/5906

eicker.news ᳇ tech newstechnews@eicker.news
2025-10-18

#AndrejKarpathy believes #AGI is still a decade away, citing the need for advancements in #continuallearning, #multimodality, and #computeruse. He argues that while the problems are solvable, they remain challenging. Karpathy also reflects on the history of AI, highlighting the impact of #deeplearning, #reinforcementlearning, and #largelanguagemodels on the field. dwarkesh.com/p/andrej-karpathy #tech #media #news

2025-10-09

Google, 웹 브라우저 조작하는 AI 모델 공개: 에이전트 경쟁 가속화

Google이 웹 브라우저를 사람처럼 조작하는 Gemini 2.5 Computer Use 모델을 공개하며 AI 에이전트 경쟁에 본격 합류했습니다. OpenAI, Anthropic과 함께 펼치는 에이전트 경쟁의 의미를 분석합니다.

aisparkup.com/posts/5465

JustGeekJustGeek
2025-10-08

🤖 Gemini 2.5 Computer Use : l’IA de Google qui sait utiliser un navigateur web

👉 justgeek.fr/gemini-2-5-compute

2025-09-27

So - question to all my #techie friends: do you have an #UPS, and if so, which one (or ones if multiple)? With installing my new system and a DAS with 43T+ of spinning rust drives, I think it's important that I have an UPS again. (I have an APC, but it hasn't been functional for a long time -- the battery died, and I've never been able to replace the battery).

I'm really interested in which one(s) you'd recommend from a Linux standpoint. I want to make certain I get one that can cleanly shutdown my system if needed.

#computerbild #computeruse #uninterruptiblepowersupply #ups

2025-08-23

OpenCUA’s open source computer-use agents rival proprietary models from OpenAI and Anthropic https://venturebeat.com/ai/opencuas-open-source-computer-use-agents-rival-proprietary-models-from-openai-and-anthropic/ #AI #agents #ComputerUse #future

Text Shot: As open source agents built on frameworks like OpenCUA become more capable, they could fundamentally evolve the relationship between knowledge workers and their computers. Wang envisions a future where proficiency in complex software becomes less important than the ability to clearly articulate goals to an AI agent.

He described two primary modes of work: “offline automation, where the agent leverages its broader software knowledge to pursue a task end-to-end,” and “online collaboration, where the agent responds in real-time and works side by side with the human, much like a colleague.” Basically, the humans will provide the strategic “what,” while increasingly sophisticated AI agents handle the operational “how.”
2025-08-23

OpenCUA’s open source computer-use agents rival proprietary models from OpenAI and Anthropic venturebeat.com/ai/opencuas-op #AI #agents #ComputerUse #future

Text Shot: As open source agents built on frameworks like OpenCUA become more capable, they could fundamentally evolve the relationship between knowledge workers and their computers. Wang envisions a future where proficiency in complex software becomes less important than the ability to clearly articulate goals to an AI agent.

He described two primary modes of work: “offline automation, where the agent leverages its broader software knowledge to pursue a task end-to-end,” and “online collaboration, where the agent responds in real-time and works side by side with the human, much like a colleague.” Basically, the humans will provide the strategic “what,” while increasingly sophisticated AI agents handle the operational “how.”
GOMOOT :mastodon:gomoot@mastodon.uno
2025-03-12
2025-01-23

ByteDance’s UI-TARS can take over your computer, outperforms GPT-4o and Claude https://venturebeat.com/ai/bytedances-ui-tars-can-take-over-your-computer-outperforms-gpt-4o-and-claude/ #AI #ComputerUse #agents

Text Shot: Researchers emphasized that it is critical that the model be able to maintain consistent goals and engage in trial and error to hypothesize, test and evaluate potential actions before completing a task. They introduced two types of data to support this: error correction and post-reflection data. For error correction, they identified mistakes and labeled corrective actions; for post-reflection, they simulated recovery steps. 

“This strategy ensures that the agent not only learns to avoid errors but also adapts dynamically when they occur,” the researchers write.

Client Info

Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst